Mass spectrometric study of the partial phase diagram of the Fe-Ge system above 1050°C

Abstract

The advantages of applying mass spectrometry to the study of phase equilibrium at high temperatures have been discussed by Gokcen et al. tal In the present study, the partial phase diagram of the Fe-Ge system in the concentration range of 4.3 to 40.9 at. pct Ge above 1050 ~ has been studied by the application of this method. Early studies on the phase diagram of the Fe-Ge system have been summarized by Shunk t2] and Kubaschewski.t3] The e phase is formed by a peritectic reaction in Shunk's summary, but this phase is formed by a peritectoid in Kubaschewski's summary. Although the phase diagram reported by l]belacke1441 was referred to in Kubaschewski's summary, the e phase shown in his paper is formed by a peritectic reaction. Recently, Enoki et al. ]SJ studied this system at temperatures below 1115 ~ and indicated that the e phase was formed by a peritectic reaction as reported by Ubelacker. The eutectic reactions reported by Shunk and Kubaschewski occurred at 1130 ~ on the other hand, the reactions reported by l]belacker and Enoki et al. occurred at 1105 ~ The solid solubility limit of Ge in a-Fe was reported as 22 at. pct Ge at 1130 ~ by Kubaschewski and 19 at. pct Ge at 1163 ~ by l]belacker. The solubility determined by Enoki et al. is 23.8 at. pct Ge at 1100 ~ and the solid solubility of Ge in a-Fe results obtained by them differ from the other literature values. A modified commercial mass spectrometer model RM-6K (Hitachi, Japan) was used in the present study. The Knudsen cell which was made of beryllia was 11 mm long and had an 8 mm inner diameter with a 1.3 mm wall thickness and a 0.5 mm orifice. The Knudsen cell was heated by the radiation from a helical coil filament of tungsten. The temperature of the Knudsen cell was measured with a Pt-Pt. 13 pct Rh thermocouple attached to the center of the bottom and calibrated against the melting point of pure nickel and copper. The alloys were prepared from 99.95 pct pure iron and 99.999 pct pure germanium. The iron and germanium were deoxidized previously in a flowing stream of H2 for three hours at 1500 ~ and for three hours at 1000 ~ respectively. The alloys (except those containing 4.3 or 7.5 at. pct Ge) were prepared in situ by melting together a total of 1.2 to 1.4 g of the component metals in each Knudsen cell. These alloys were prepared by fusing the pure component metals under a hydrogen atmosphere in a high frequency induction furnace, because the liquidus temperatures of these alloys are high. A small amount of beryllium (0.3 to 1.0 at. pct) was added to the sample